Overview: Too many depressed patients are improved but not well after "successful" antidepressant treatment, including cognitive therapy (CT). We developed continuation phase cognitive therapy (C-CT) and have shown that C-CT reduces relapse. Now we propose to raise the standard for evaluating C-CT by examining its long-term effects on psychiatric, psychosocial, and economic outcomes. Background: In the last project period, we tested the efficacy of C-CT in reducing relapse in outpatients who presented with recurrent MDD and responded to A-CT (i.e., 20 sessions of cognitive therapy). Responders were randomized to an 8 month (10 session) experimental phase with two cells: (a) C-CT or (b) CT-free evaluation sessions (control). As predicted, C-CT reduced relapse (i.e., DSM-IV MDD) significantly more than control over 8 months, according to an evaluator blind to treatment assignment. New Aims: Now we propose to compare C-CT to control in improving long-term psychiatric, psychosocial. and economic outcomes over 24 months in these outpatients who presented with recurrent major depressive disorder (MDD) and then responded to acute phase cognitive therapy (A-CT). Through new analyses, we will evaluate the effects of C-CT versus control in restoring cognitive, interpersonal, social, and work functioning, as well as in reducing the cost of care for both mental and general health. We propose five sets of hypotheses to scrutinize long-term effects of C-CT on this broad range of outcomes, to identify predictors of outcomes and to discover mechanisms of change relevant to building and revising our model describing how C-CT prevents relapse.